Energy Information Administration Home Page Welcome to the Energy Information Administration's Energy Efficiency Web Site.  If you need assistance in viewing this page, please call (202) 586-8800

  Home > Energy Users > Energy Efficiency Page > Report: Measuring Energy Efficiency in the U.S. Economy

Measuring Energy Efficiency In The United
States' Economy: A Beginning

Measuring Energy Efficiency in the United States' Economy (file size .94 MB) pages: 105. To View and/or Print Report (requires Adobe Acrobat Reader) - Download Adobe Acrobat Reader.

Executive Summary


Defining Energy Efficiency and Its Measurement

Residential Sector

Commercial Buildings Sector

Transportation Sector

Industrial Sector

United States Economy

Future Directions

Appendix:  Methodology




October 1995

Energy Information Administration

Office of Energy Markets and End Use

U.S. Department of Energy

Washington, DC 20585

Executive Summary

Energy efficiency is a vital component of the Nation's energy strategy. One of the Department of Energy's missions are to promote energy efficiency to help the Nation manage its energy resources. The ability to define and measure energy efficiency is essential to this objective. In the absence of consistent defensible measures, energy efficiency is a vague, subjective concept that engenders directionless speculation and confusion rather than insightful analysis.

The task of defining and measuring energy efficiency and creating statistical measures as descriptors is a daunting one. This publication is not a final product, but is EIA's first attempt to define and measure energy efficiency in a systematic and robust manner for each of the sectors and the United States'    economy as a whole. In this process, EIA has relied on discussions, customer reviews, in-house reviews, and seminars that have focused on energy efficiency in each of the sectors. EIA solicits the continued participation of its customers in further refining this work.

The term energy intensity is used to describe a statistical ratio estimator of energy use relative to energy service demand. This estimator is often used as a measurement indicator of energy efficiency; and in fact, the concepts of intensity and efficiency are sometimes used interchangeably. Intensity measures must be based on available data, and so may not reflect factors that might, if measurable, allow energy intensity to approximate energy efficiency more precisely; however, trends in the energy-intensity indicators are generally suggestive of trends in energy efficiency. Trends in energy intensity can be influenced by factors other than energy efficiency. As a practical matter, it is virtually impossible to remove, or even to consider, all of the behavioral or structural factors that would be necessary to obtain a pure measurement of energy efficiency, however broadly energy efficiency may be defined.

In this report, the focus is on the measurement of energy intensity. Wherever possible, available data are used to remove from the energy-intensity indicator influencing factors such as weather, capacity, and inventory changes that are commonly viewed as not related to changes in energy efficiency. For each sector, the number of different energy-intensity indicators presented is limited by the data available.

Each chapter covering an economic sector contains discussions of:

  • Trends in energy consumption for the economic sector and the factors that influence changes in energy consumption
  • Description of the data used to develop indicators of energy-intensity
  • Demand indicators that influence the amount of energy consumed as well as their changes over time
  • Energy-intensity indicators that compute the ratio of energy consumption to a unit of measurement, for example, Btu per square foot of floorspace, or Btu per household
  • Strengths and limitations of the energy-intensity indicators such as whether data adequately cover the sector, whether data are of good quality and timely, or whether demand indicators characterize energy demand for all required services.

Each energy-intensity indicator is examined over two intervals of time: one beginning and ending with years of economic growth and one beginning with growth and ending with recession or in the case of the commercial building sector, beginning with recession and ending with recovery. The following is a summary of findings for each economic sector and the economy as a whole. A graphical presentation (Figures ES1 a-l) of these findings follows the sector discussions.

Residential Sector

In the residential sector, four different energy-intensity indicators were developed using data from EIA's Residential Energy Consumption Survey (RECS) (Figures ES1a and ES1b). Each of the indicators was adjusted for the influence of weather. These adjustments led either to lower reductions in the energy-intensity indicators during the growth/growth interval (1984 to 1987) or to actual increases in energy-intensity indicators during the growth/recession interval (1987 to 1990). Adjustments for the effects of weather appear to be valid. Without these adjustments, efficiency gains in the residential sector would be overestimated. Real efficiency gains appeared to have been made in the residential sector, since most intensity indicators displayed reductions; however, the magnitude of the gain cannot be determined from available data. Other structural influences, such as changes in disposable income and the mix of housing types and energy sources, may play a significant role, and the methodology to account for these and other behavioral and structural effects needs to be developed.

Commercial Building Sector

Energy-intensity indicators in the commercial buildings sector increased during the growth/growth interval (1986 to 1989), with the exception of the Btu per square-foot-hour indicator. Five energy-intensity indicators were developed, four with data from EIA's Commercial Buildings Energy Consumption Survey (CBECS) (Figures ES1c and ES1d). All five energy-intensity indicators decreased during the recession/recovery interval (1989 to 1992). Adjustments for weather and vacancy effects, for the most part, dampened the increases in energy-intensity indicators during the growth/growth interval and reduced the decrease in the energy-intensity indicators during the recession/recovery interval.

It is difficult to ascertain if real efficiency gains were made in the commercial buildings sector. Only one energy-intensity indicator was consistent over both the growth/growth and recession/recovery intervals, Btu per square-foot-hour. This indicator encompasses both the size of the building and hours of usage, two major factors affecting energy usage. This indicator does show possible efficiency gains over both time intervals. More work needs to be done to capture other structural and behavioral effects that may influence changes in energy usage in this sector.

Transportation Sector

The transportation sector is one of the most difficult and complex sectors in which to determine whether efficiency gains have been made. Data are very sparse for this sector. No one survey covers the entire transportation sector in the same way the residential or commercial sector is covered, either for passenger transportation or freight. EIA fields the Residential Transportation Energy Consumption Survey (RTECS), where the fuel consumption and expenditures are based on estimates using vehicle efficiencies from the U.S. Environmental Protection Agency (EPA). RTECS does not collect data to determine passenger miles traveled, and it dropped consumption data based on fuel logs in 1985. Other passenger mode and freight mode data are collected from other sources such as Transportation Statistics: Annual Report 1994, published by the Bureau of Transportation Statistics, and the Transportation Energy Data Book: Edition 14, developed by the Oak Ridge National Laboratory for the Department of Energy. The National Personal Transportation Survey is fielded by the U.S. Department of Transportation only once every 7 years. Therefore, any energy-intensity indicators developed for the passenger, freight, or the transportation sector as a whole, should be viewed with caution.

For the passenger transportation sector, the best energy-intensity indicator seems to be thousand Btu per passenger mile. This intensity indicator accounts for the differences in passenger loads between the automobile and mass transit. For all passenger transportation modes, the energy-intensity indicator appeared to be on the decline or at least flat over the growth/growth interval (1985 to 1988) (Figures ES1e and ES1f). Energy-intensity indicator for mass transit seemed to be on the rise over the growth/recession interval (1988 to 1991) as passenger loads dipped. In general, the passenger transportation sector appears to be less energy intensive, a finding that implies that it may have become more energy efficient over both time intervals.

Changes in the energy-intensity indicator, energy per ton mile, for the freight transportation sector produced mixed results over the different modes (Figures ES1g and ES1h). It is difficult to determine whether a particular mode in the sector became more or less intensive over both time intervals. The growth/growth interval (1985 to 1988) shows a large percent increase in the energy-intensity indicator (54.5 percent). This distortion is due to the fact that small package data collection for air carriers, such as United Parcel Service, did not begin until 1986. However, for the growth/recession interval (1988 to 1991), the air-carrier mode seems to have become slightly more energy intensive. Rail is the only mode with energy-intensity indicators declining over both intervals, implying decreases in energy-intensity and increases in energy efficiency. Again, effects other than efficiency may influence these declines.

Industrial Sector

Even more troublesome data problems exist in the industrial sector. Data for the agriculture, mining, and forestry sectors are suitable only for rudimentary analysis, and they are not presented in this publication.

The Manufacturing Energy Consumption Survey (MECS), fielded by EIA, is the most comprehensive source of manufacturing energy consumption data, although several other sources exist. Seven different energy-intensity indicators are presented for the manufacturing sector as a whole, and for each of three consuming groups: High-Energy Consumers, High-Value added Consumers, and Low-Energy Consumers.

Improvements in energy efficiency in the manufacturing sector seemed to be slight, even when the effects of changes in inventory and capacity were considered (Figures ES1i and ES1j). After adjustments for these two effects, the indicator (energy per adjusted value of production) actually increased 5.0 percent over the growth/growth interval (1985 to 1988) and fell by 6.6 percent over the growth/recession interval (1988 to 1991), leaving only a 1.6-percent decrease in this energy-intensity indicator over both intervals of time.


An economy-wide energy-intensity indicator that is consistent over all sectors does not exist, except for one that uses an economic-demand indicator such as Gross Domestic Purchases, Gross Domestic Product (GDP) or population. In this report, the approach was to build a composite indicator from intensity changes and shares for each sector of the economy. EIA consumption surveys are fielded in rotating years, e.g., MECS in 1985, CBECS in 1986, and RECS in 1987, and a methodology to compensate for this obstacle was developed. Although far from perfect, the composite indicators represent a beginning in the development of an economy-wide, energy-intensity indicator.

All three of the composite indicators, as well as the energy per Gross Domestic Purchases presented in Figure 7.1 and discussed in Chapter 7, showed little or no reductions in energy intensity over both time intervals, indicating very little efficiency gains (Figures ES1k and ES1l). One exception was site energy per capita during the growth/growth interval (1985 to 1988), which increased by 6.1 percent and decreased by only 2.6 percent during the growth/recession interval (1988 to 1991). Such a finding may indicate the influence of structural and behavioral effects rather than true gains/losses in efficiency. More research is needed to refine these composites.

This report is intended to stimulate discussion and invite reader feedback and future participation.

This is indeed a beginning task. EIA, with the assistance of its customers, needs to refine the energy-intensity indicators as presented. Refining these indicators will require more work, such as:

  • Development of indicators of energy-intensity--some demand indicators, like floorspace for commercial and residential sectors, do not completely capture the energy-demand profile. Other demand indicators or combinations of indicators need to be defined.
  • Identification and analysis of structural and behavioral effects. Energy estimates for each sector were adjusted for factors such as weather, occupancy, inventory changes, and capacity utilization; however, other effects may influence energy efficiency as well. Methodology is needed to adjust for other structural and behavioral effects.

Customer participation in the form of report feedback and workshop participation will assist in obtaining the most feasible and robust indicators within the limitation of the data and resources available.

Back to Reports Reports

Energy Efficiency Measurement:   Discussion

blue_next.gif (2052 bytes) Introduction


Specific questions on this topic may be directed to:
Stephanie Battles
(Phone: (202) 586-7237)
FAX: (202) 586-0018

Contact Us


File Last Modified: October 17, 1999